[libstdc++] Make use of runtime demangler

Franz Sirl Franz.Sirl-kernel@lauterbach.com
Tue Apr 30 17:12:00 GMT 2002


On Wednesday 01 May 2002 01:41, Phil Edwards wrote:
> On Tue, Apr 30, 2002 at 12:20:42AM +0200, Franz Sirl wrote:
> > > So, before we start reintroducing any of that, I have to wonder whether
> > > the same problems -- presumably problems with shadow headers -- will
> > > still occur. Probably not a fair question, since so much of our header
> > > setup has changed in the last 18 months.
> >
> > But why would one remove the whole of $(DEFS) just for the sake of
> > removing a single include?
>
> I have no idea.  But I don't recall what problems we were experiencing at
> the time.  (It was 18-odd months ago, I can't even remember when to go to
> lunch from day to day.)

Heh :-)

> > Simply setting $(DEFS) as required in Makefile.am seems like
> > a lot simpler approach to me. Do you want to go with such an solution? I
> > would additionally add
> >
> > # we need DEFS without -I$(srcdir)
> > DEFS = @DEFS@ -I. -I..
> >
> > to Makefile.am in my patch then.
>
> Actually, we shouldn't even need the -I switches, since we're not adding
> them back in anywhere else, I /think/.  There are probably some multilib
> issues here.

We need at least -I.. to get at config.h and I don't see any multilib issues 
here, since -I.. will be $target_alias/$multilibdir/libstdc++-v3/

> I stuck the DEFS definition as shown below; it seemed to make the most
> sense there.  Would you be willing to add that change to your patch and
> retest?  Probably a retest isn't necessary, but I'm once-bitten-twice-shy
> when playing with variables that can be automatically set by automake.
> I think the patch can go in after that.

I'll retest with the -I.. added, seems like I'm doing nothing but bootstraps 
the whole day currently :-).

Franz.



More information about the Gcc-patches mailing list